Addition of Lateral Extra-Articular Tenodesis to Primary Anterior Cruciate Ligament Reconstruction in Competitive Athletes with High-Grade Pivot-Shift Is Associated with Lower Graft Failure and Faster Return to Sport: A Propensity Score-Matched Multicentre Cohort Study

Abstract

Aim of the Study: To determine whether adding a lateral extra-articular tenodesis (LET) to primary anterior cruciate ligament reconstruction (ACLR) lowers graft-failure risk and improves functional recovery in competitive athletes with high-grade pivot-shift. Methods: Multicentre retrospective cohort with 1:1 propensity-score matching (age, sex, sport, graft, centre). Competitive athletes with pivot-shift grade ≥ 2 who underwent primary ACLR with hamstring or bone–patellar tendon–bone (BPTB) autografts (2018–2024) were eligible. The primary outcome was graft failure within 24 months (composite of revision ACLR, symptomatic rotatory laxity with pivot-shift ≥ 2 plus KT-1000 > 5 mm, or MRI-confirmed rupture). Time-to-event was summarised with Kaplan–Meier (KM) curves and log-rank tests. Secondary outcomes included residual rotatory laxity and functional performance (single-leg hop, side hop, Y-Balance) analysed as the proportion achieving Limb Symmetry Index ≥ 90% at 6 and 24 months and as continuous LSI means. Two-sided α = 0.05; secondary outcomes were prespecified without multiplicity adjustment. Results: Of 1368 ACL reconstructions screened, 97 eligible athletes were identified; 92 were analysed after matching (46 isolated ACLR; 46 ACLR + LET; mean follow-up 30.0 ± 4.2 months). KM survival at 24 months was 95.7% after ACLR + LET versus 82.6% after isolated ACLR (log-rank p = 0.046). The absolute risk reduction was 13.0% (Number Needed to Treat 8; 95% CI 4→∞). In graft-type subgroups, failures were 6/32 vs. 1/30 for hamstring and 2/14 vs. 1/16 for BPTB (ACLR vs. ACLR + LET, respectively); there was no evidence of interaction (Breslow–Day p = 0.56). At 6 months, a higher proportion of ACLR + LET athletes achieved LSI ≥ 90% across tests—single-leg hop 77.8% vs. 40.9% (p = 0.0005), side hop 62.2% vs. 34.9% (p = 0.012), Y-Balance 84.4% vs. 59.1% (p = 0.010), with a larger mean LSI (between-group differences +8.2 to +9.1, all p < 0.001). By 24 months, threshold attainment largely converged (all p ≥ 0.06), while mean LSI differences persisted but were smaller (+3.9 to +4.9, all p ≤ 0.001). Conclusion: In competitive athletes with high-grade pivot-shift undergoing accelerated, criteria-based rehabilitation, adding LET to primary ACLR was associated with lower graft-failure risk and earlier functional symmetry, with consistent effects across hamstring and BPTB autografts. Given the observational design, causal inference is limited; confirmation in randomized and longer-term studies is warranted.

1. Introduction

Anterior cruciate ligament (ACL) rupture is a potentially career-ending injury for athletes in pivoting sports, with >150,000 reconstructions performed annually in the United States alone [1]. Despite advances in arthroscopic technique and rehabilitation, graft-failure rates of 3–11% and symptomatic residual rotatory laxity approaching 30% persist [2,3]. Residual rotatory laxity may persist after anatomic ACLR because the anterolateral complex is only partially restored, the graft may elongate during early healing, and tunnel malpositioning can leave unchecked tibial internal rotation [4]. High-grade (grade 2–3) pivot-shift pre-operatively, young age, high-level pivoting sport participation and generalised laxity are the strongest independent predictors of failure [5,6,7]. Anatomical and biomechanical work has highlighted the contributions of the anterolateral structures, iliotibial band fibres (Kaplan), and anterolateral ligament (ALL) to controlling internal tibial rotation [8,9]. These insights revived interest in extra-articular augmentation, particularly the modified Lemaire LET, which unloads the intra-articular graft during pivoting motions [4,10]. Modern randomised trials (STABILITY I and II) demonstrated that adding LET to hamstring ACLR in young, high-risk patients reduced graft failure by 40–60% [11,12]. Conversely, other series reported absolute risk reductions of only 2–4% and little change in patient-reported outcomes and questioned routine use owing to perceived morbidity [13].

Propensity-score methodology offers a pragmatic alternative to randomisation in surgical research, balancing measured covariates and reducing selection bias [14]. To our knowledge, no multicentre propensity-matched study has evaluated LET specifically in competitive European athletes undergoing accelerated, criteria-based rehabilitation incorporating neuromuscular electrical stimulation (NMES) and blood-flow-restriction training (BFRT); both interventions have been shown to mitigate early quadriceps inhibition and atrophy [15,16]. However, no study has examined whether LET remains advantageous when modern, criterion-based accelerated rehabilitation is applied to European athletes who must resume competition rapidly. Unlike STABILITY I, this multicentre propensity-matched cohort includes both hamstring and BPTB autografts and applies a criteria-based accelerated rehabilitation (including NMES and BFRT) among competitive European athletes. Propensity-score matching was selected to address key baseline confounders that are not reliably modelled by linear covariate adjustment, including age, sex, graft type (hamstring vs. BPTB), sport and contact level, and centre (proxy for surgeon/rehabilitation variability). Focusing on European competitive athletes minimises contextual heterogeneity (league calendars, return-to-play regulations, and sport mix with a predominance of field sports), yielding a more homogeneous exposure–outcome environment than internationally pooled cohorts. A multicentre design enhances external validity and dilutes single-surgeon effects, while matching on centre mitigates systematic differences in surgical technique and perioperative care across hospitals.

Graft survival, residual instability, and functional recovery were compared between isolated ACLR and ACLR combined with LET in a matched cohort from three high-volume sports medicine centres. It was hypothesized that LET would (1) reduce the risk of graft failure, (2) decrease residual pivot-shift grading, and (3) expedite return to sport without increasing complication rates.

2. Methods

2.1. Study Design and Reporting

A multicentre retrospective cohort study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. The protocol was approved by the regional ethics committee from Università Degli Studi del Molise (No. 23/2022). All procedures complied with the Declaration of Helsinki.

2.2. Setting and Participants

Electronic surgical logs from the three participating centres were screened for primary ACLR (January 2018–February 2024). Inclusion criteria: (i) competitive athlete (≥4 sessions/week and participation in organised competition), (ii) age 16–35 years, (iii) pre-operative pivot-shift grade 2–3 confirmed under anaesthesia, (iv) hamstring or bone–patellar tendon–bone autograft. Exclusions: multiligament reconstruction, previous ipsilateral knee surgery, contralateral ACL injury, significant osteochondral defect (area: >2 cm²), varus/valgus malalignment >5°, presence of meniscal tears classified as Bad and Ugly by Simonetta et al. [17], skeletal immaturity, incomplete data or <12 months follow-up. The decision to add LET followed a predefined algorithm: age < 25 years, pivot-shift grade 3, participation in Level I pivoting sport, generalised ligamentous laxity or genu recurvatum > 10°. No surgeon preference or patient request overrode these criteria (Figure 1).

2.3. Propensity-Score Matching

To minimise allocation bias, a 1:1 nearest-neighbour algorithm (calliper 0.1, no replacement) matched athletes undergoing ACLR + LET to isolated ACLR based on age, sex, graft type, sport (contact vs. non-contact) and centre. Balance was assessed with standardised mean differences; values < 0.10 were accepted as adequate (Table S1).

2.4. Surgical Techniques

All procedures were performed by fellowship-trained sports surgeons (>100 ACLRs/year). In both groups, a standard anteromedial portal, anatomic single-bundle technique was used. Hamstring autografts (diameter ≥ 8 mm) were fixed with suspensory cortical buttons femorally and bioabsorbable interference screws tibially; BPTB grafts received metal interference screws at both tunnels. Graft selection (hamstring vs. BPTB) followed predefined criteria (sport demands, graft diameter, prior anterior knee pain) agreed across centres.

LET procedure: A 1 × 10 cm strip of iliotibial band was harvested, passed deep to the lateral collateral ligament and fixed proximal to the lateral epicondyle with a 6 mm bio-composite interference screw at 30° knee flexion, neutral rotation, with 20 N tension (modified Lemaire). Indications for LET mirrored those in STABILITY I: age < 25 years, grade 3 pivot-shift, high-risk sport (Level I), generalised laxity or genu recurvatum > 10°.

2.5. Rehabilitation

Phase 1 (weeks 0–4): immediate full extension, partial weight-bearing with crutches (first 7 days), NMES twice daily and BFRT (3 sets, 30-15-15 reps, 80 mmHg) from day 3. Phase 2 (weeks 5–12): progressive closed-kinetic-chain strengthening, stationary cycling, and pool running; open-chain quadriceps from week 8. Phase 3 (weeks 12–18): plyometric and agility drills; jogging initiated when limb-symmetry-index (LSI) quadriceps torque at 60°/s ≥ 70%. Phase 4 (criteria-based RTS): unrestricted sport allowed when (i) LSI quadriceps and hamstrings ≥ 90%, (ii) hop-test battery average LSI ≥ 90%, (iii) IKDC-Subjective ≥ 80 and (iv) ACL-RSI ≥ 70.

2.6. Outcomes

Primary outcome: Graft failure within 24 months, defined as (a) revision ACLR, (b) symptomatic laxity with positive pivot-shift grade ≥ 2 plus side-to-side KT-1000 difference > 5 mm or (c) clinician-diagnosed graft rupture on MRI.

Secondary outcomes: (1) residual pivot-shift grade ≥ 1, (2) IKDC-Subjective (0–100), (3) ACL-RSI (0–100), (4) time to RTS (months from surgery to first full competitive match), (5) complications (Clavien–Dindo), (6) patient-reported satisfaction (5-point Likert). All KT-1000 measurements were obtained by a single independent physiotherapist who was not involved in the index surgery and was blinded to treatment allocation at 6, 12 and 24 months. Residual pivot-shift (secondary) was analysed independently of the composite failure definition; failure required grade ≥ 2 plus anterior laxity (KT-1000 > 5 mm) to avoid circularity.

2.7. Sample Size and Detectable Difference

The available matched sample (n = 92) afforded 80% power (two-sided α 0.05) to detect a 13% absolute difference in graft failure (baseline 17%) using χ² testing.

2.8. Statistical Analysis

Continuous variables were tested for normality (Shapiro–Wilk). Means (±SD) or medians (IQR) are provided. Categorical variables are presented as n (%). Group comparisons employed Welch two-sample t-tests (continuous) and Fisher’s exact tests (categorical). Multivariable logistic regression estimated aORs for graft failure and residual pivot-shift, adjusting for meniscal procedure, graft diameter and centre. Linear mixed models examined longitudinal PRO scores. Kaplan–Meier curves described graft survival; Greenwood SEs generated 95% CIs. The significance threshold was set at α 0.05. Missing data (<2%) were handled with multiple imputation. Secondary outcomes were prespecified; p-values are presented without multiplicity adjustment. A sensitivity analysis applying Holm–Bonferroni control yielded the same conclusions for outcomes with p < 0.01. Pre-specified subgroup analyses assessed effect modification by autograft type (hamstring vs. BPTB) using an interaction term (LET×graft) in logistic models; the Breslow–Day test was used to evaluate homogeneity of odds ratios across strata. Absolute risk reduction (ARR) used Newcombe–Wilson CIs; NNT = 1/ARR with Altman’s approach for confidence limits (i.e., upper bound ∞ when the ARR 95% CI encompassed zero).

3. Results

From 1368 primary ACLRs screened, 97 eligible competitive athletes were identified; 92 were included after 1:1 propensity-score matching (46 isolated ACLR; 46 ACLR + LET). All post-match standardized mean differences were <0.10, indicating excellent covariate balance in line with recommended thresholds for matched studies. The groups were identical for age (24.8 ± 4.3 vs. 24.6 ± 4.6 years), sex distribution (≈ 68% men), graft type, high-risk sport participation and concomitant meniscal repair (

Table 1. Baseline characteristics of the matched cohort. Baseline characteristics after 1:1 propensity-score matching (n = 92). Values are mean ± SD or n (%). Standardized mean differences (SMD) are shown for covariates used in matching; all SMD < 0.10.

Characteristic	ACLR (n = 46)	ACLR + LET (n = 46)	SMD (Post-Match)
Age, years (mean ± SD)	24.8 ± 4.3	24.6 ± 4.6	0.045
Male sex, n (%)	32 (69.6)	31 (67.4)	0.047
Hamstring graft, n (%)	28 (60.9)	29 (63.0)	0.043
High-risk sport (Level I), n (%)	38 (82.6)	39 (84.8)	0.060
Concomitant meniscal repair, n (%)	12 (26.1)	13 (28.3)	0.049
Follow-up, months (mean ± SD)	29.7 ± 5.2	30.1 ± 5.0	—

).

3.1. Graft Survival and Failure Analysis

Over a mean follow-up 30.0 ± 4.2 months, graft failure was recorded in eight isolated reconstructions (17.4%) but in only two augmented reconstructions (4.3%). Six of the ten graft failures were non-contact pivots in football; two occurred during alpine skiing and two during basketball landings. Multivariable logistic regression, adjusting for graft diameter and meniscal procedure, confirmed a significant association between LET and lower graft failure (adjusted OR 0.21, 95% CI 0.04–0.98; p = 0.047). The Kaplan–Meier curve demonstrated 95.7% graft survival at 24 months for ACLR + LET versus 82.6% for isolated ACLR, yielding an absolute risk reduction of 13.0% and a number-needed-to-treat of 8. Absolute risk reduction (ACLR–LET): HS 15.4% (95% CI −7.8 to +34.7; NNT ≈ 7, CI 3→∞), BPTB 8.1% (95% CI −24.3 to +38.8; NNT ≈ 13, CI 3→∞) (Figure 2).

3.2. Graft-Type Subgroup Analysis

When stratified by autograft, failure proportions were 6/32 (18.8%) after isolated ACLR versus 1/30 (3.3%) after ACLR + LET among hamstring reconstructions, and 2/14 (14.3%) versus 1/16 (6.3%) among BPTB reconstructions. Within hamstring ACLR, the odds of graft failure were lower with LET (odds ratio 0.26, 95% CI 0.04–1.64; Fisher p = 0.105); a similar trend was observed for BPTB (odds ratio 0.74, 95% CI 0.09–6.43; Fisher p = 0.586). The LET×graft interaction was not significant (Breslow–Day p = 0.56), indicating no statistical evidence that the association between LET and failure differed by graft type. Consistent with the primary analysis, the log-rank test for overall graft survival favoured ACLR + LET (p = 0.046) (Table S2).

3.3. Residual Rotatory Stability

Clinical assessment at final review found a grade-1 or higher pivot-shift in 17 isolated ACLRs (37.0%) compared with six ACLR + LET knees (13.0%). After adjustment, LET reduced the odds of residual rotatory laxity by 76% (adjusted OR 0.24, 95% CI 0.09–0.65; p = 0.004).

3.4. Patient-Reported Outcome Measures

Patient-reported outcomes improved in both cohorts but consistently favoured the augmented reconstruction. At 24 months, the mean IKDC-Subjective score was 6.2 points higher after LET (86.4 ± 6.1 vs. 80.2 ± 7.4; 95% CI 2.9–9.5; p < 0.001). Psychological readiness mirrored this pattern: the ACL-RSI scale, which captures emotion, confidence and risk appraisal during return to sport [18], reached 78.5 ± 9.2 after LET versus 69.1 ± 10.4 after isolated ACLR (difference 9.4 points; p = 0.001). Linear mixed-effects modelling confirmed that the LET advantage became apparent as early as six months and persisted throughout follow-up (p interaction = 0.02).

3.5. Functional Performance Tests

At 6 months, a greater proportion of athletes achieved LSI ≥ 90% after ACLR + LET across all tests: single-leg hop 77.8% (35/45) vs. 40.9% (18/44) (risk difference +36.9%, 95% CI +8.1 to +59.8; Fisher p = 0.0005); side hop 62.2% (28/45) vs. 34.9% (15/43) (RD +27.3%, 95% CI −2.2 to +52.5; p = 0.0118); Y-Balance composite 84.4% (38/45) vs. 59.1% (26/44) (RD +25.4%, 95% CI −1.1 to +47.8; p = 0.0098). At 24 months, threshold attainment converged between groups: single-leg hop 95.6% vs. 83.7% (RD +11.8%, 95% CI −6.7 to +28.7; p = 0.086), side hop 93.3% vs. 78.6% (RD +14.8%, 95% CI −6.2 to +33.6; p = 0.063), Y-Balance 97.8% vs. 90.5% (RD +7.3%, 95% CI −7.8 to +21.7; p = 0.192). As continuous LSI, ACLR + LET demonstrated higher symmetry at both time points: at 6 months, mean between-group differences (LET–ACLR) were +8.2 (95% CI +5.10 to +11.30) for single-leg hop, +9.1 (95% CI +5.57 to +12.63) for side hop, and +7.4 (95% CI +4.90 to +9.90) for Y-Balance (all p < 0.001). At 24 months, differences persisted but were smaller: +4.1 (95% CI +1.76 to +6.44), +4.9 (95% CI +2.03 to +7.77), and +3.9 (95% CI +1.76 to +6.04) respectively (all p ≤ 0.001) (

Table 2. Functional performance; proportions with LSI ≥ 90%.

Test	Time	ACLR ≥ 90%	ACLR + LET ≥ 90%	Risk Difference (LET–ACLR)	95% CI	p (Fisher)
Single-Leg Hop	6 mo	18/44 (40.9%)	35/45 (77.8%)	+36.9%	+8.1 to +59.8	0.0005
Side Hop	6 mo	15/43 (34.9%)	28/45 (62.2%)	+27.3%	–2.2 to +52.5	0.0118
Y-Balance	6 mo	26/44 (59.1%)	38/45 (84.4%)	+25.4%	–1.1 to +47.8	0.0098
Single-Leg Hop	24 mo	36/43 (83.7%)	43/45 (95.6%)	+11.8%	–6.7 to +28.7	0.0860
Side Hop	24 mo	33/42 (78.6%)	42/45 (93.3%)	+14.8%	–6.2 to +33.6	0.0631
Y-Balance	24 mo	38/42 (90.5%)	44/45 (97.8%)	+7.3%	–7.8 to +21.7	0.1924

).

3.6. Return-to-Sport Timeline

Athletes rehabilitated with an identical criteria-based protocol, including early neuromuscular electrical stimulation (NMES) and blood-flow-restriction training, achieved benchmark strength and hop symmetry sooner when LET had been added. The mean time to unrestricted sport was 5.3 ± 1.1 months after LET versus 7.2 ± 1.3 months after isolated ACLR (mean difference –1.9 months; p < 0.001). At the eight-month mark, 83% of LET athletes and 54% of isolated ACLR athletes had resumed full competition (risk ratio 1.54, 95% CI 1.15–2.06). No second ACL injuries were identified between 24 and 30 months; nevertheless, surveillance beyond 3 years is warranted to assess long-term reinjury after accelerated RTS.

3.7. Complications and Satisfaction

Safety profiles were comparable. No deep infections, neurovascular injuries or symptomatic over-constraint were recorded. The LET cohort sustained one superficial lateral incision infection, which was resolved with a week of oral flucloxacillin and left no sequelae. Overall satisfaction was high but higher after augmentation: 78% of LET athletes rated their outcome “very satisfactory” versus 59% after isolated ACLR (p = 0.048). Emerging evidence suggests that perceived stability and confidence, both enhanced by LET, are pivotal drivers of postoperative satisfaction (

Table 3. Clinical outcomes at 24 months.

Outcome	ACLR	ACLR + LET	Effect (95% CI)	p
Graft failure n (%)	8 (17.4)	2 (4.3)	aOR 0.21 (0.04–0.98)	0.047
Residual pivot-shift ≥ 1 n (%)	17 (37.0)	6 (13.0)	aOR 0.24 (0.09–0.65)	0.004
IKDC-S, mean ± SD	80.2 ± 7.4	86.4 ± 6.1	MD 6.2 (2.9–9.5)	<0.001
ACL-RSI, mean ± SD	69.1 ± 10.4	78.5 ± 9.2	MD 9.4 (4.1–14.7)	0.001
Time-to-RTS, months, mean ± SD	7.2 ± 1.3	5.3 ± 1.1	MD –1.9 (–2.4 to –1.4)	<0.001
Complications, n	0	1 (superficial infection)	—	0.31

).

4. Discussion

Augmenting anatomic hamstring/BPTB ACLR with a modified Lemaire LET in young competitive athletes with pronounced pre-operative pivot-shift was associated with a 79% relative reduction in graft failure, halved residual rotatory laxity and accelerated RTS by nearly two months without clinically important morbidity. These findings are consistent with the multicentre STABILITY I RCT, in which LET reduced failure rates from 11% to 4% in high-risk patients [11]. However, STABILITY utilised allograft augmentation in some cases and mandated a 9 month RTS restriction. The cohort in the current study utilized autografts exclusively and followed an evidence-based, criteria-driven RTS algorithm, while achieving graft survival rates comparable to those reported in STABILITY I, suggesting that the protective effect of LET persists under accelerated protocols. Ripoll et al. reported a 61% relative reduction in failure with LET in pivoting athletes <20 years [19]. Conversely, Getgood et al. found no difference in PROs by 24 months, although their study was under-powered and included recreational populations > 35 years [20]. The present investigation, the largest propensity-matched European series to date, strengthens the external validity of LET benefits within modern rehabilitation paradigms. LET outcomes remain heterogeneous across studies, likely reflecting differences in age, graft-type and rehabilitation pace; the present cohort extends the evidence to accelerated protocols in professional sport yet underscores that benefits may be smaller in older or recreational populations. LET benefits were confirmed within a propensity-matched, accelerated-rehabilitation context and across common autograft types (hamstring and BPTB) in competitive athletes.

4.1. Mechanistic Considerations

Cadaveric and finite-element studies affirm LET off-loads the intra-articular graft by ~40% during pivoting loads and limits internal rotation by 3–5° [21,22]. The current clinical data, significant reduction in residual pivot-shift and higher ACL-RSI, support these biomechanical models. Greater perceived stability likely explains the observed expedient RTS, echoing psychometric research identifying confidence as a dominant RTS predictor [23]. Collectively, these findings indicate earlier and greater functional symmetry after ACLR + LET, with near-universal threshold attainment in both groups by 24 months.

4.2. Rehabilitation Synergy

Integrating NMES and blood-flow-restriction training (BFRT) into early rehabilitation mitigated quadriceps inhibition and atrophy, reflected in low complication rates and robust LSI values. Meta-analyses confirm NMES enhances early strength recovery post-ACLR [15], while BFRT achieves hypertrophy at low load, reducing graft strain [16]. Because both groups followed the same protocol, between-group differences are unlikely to be attributable to rehabilitation content; any earlier functional loading should be interpreted as hypothesis-generating. The strength gains observed may, in part, reflect the proven utility of NMES in attenuating postoperative quadriceps inhibition [24]. Notably, returning before nine months has been associated with a seven-fold increase in second-injury risk in young populations [25], underscoring the value of the enhanced stability conferred by LET for those who must resume sport sooner. Clinicians should still respect biological graft maturation and adhere to objective RTS criteria to avoid premature return, a recognised reinjury risk [25].

4.3. Clinical Implications

Selective LET should be considered in athletes displaying ≥grade 2 pivot-shift, generalised laxity or participation in Level I pivot-contact sports. The NNT = 8 sits within acceptable surgical thresholds, especially when weighed against revision costs (financial and chondral) and time-loss from sport. While standardized tensioning was used, multicentre data show up to 3% risk of over-constraint or lateral tenderness even with experienced surgeons; strict attention to the femoral insertion point and 20 N tension is therefore essential [7]. Surgeons must master the technique to avoid over-constraint, lateral tenderness or tunnel convergence, complications scarcely observed here owing to standardised graft tensioning and fixation angles. This observation is clinically important given evidence linking high pivot-shift grades to subsequent graft rupture and cartilage overload [6].

4.4. Strengths and Limitations

Strengths of this study include its multicentre design, stringent matching, homogeneous high-level athletic cohort, unified rehabilitation, prospectively collected outcomes and >80% power for the primary endpoint.

Limitations: the retrospective nature of this study precludes unmeasured confounding (e.g., tibial slope); matching cannot account for surgeon preference nuances. Follow-up averaged 30 months; longer observation is required to evaluate osteoarthritis progression. Pivot-shift grading is examiner-dependent; the use of video-based calibration sessions and blinded dual assessment mitigated this. Many cases have not been taken into consideration due to the concomitant presence of unstable meniscal lesions, such as the frequent RAMP lesions, which alter the evaluation of the single procedure [26]. Early RTS (<9 months) may increase the risk of subsequent reinjury; however, the follow-up period may have been insufficient to detect such events. Although outcome assessors were independent and intended to be blinded, surgical scars could have partly unblinded allocation. Results may not generalise to recreational patients or those receiving allografts. Propensity score methods balance measured covariates but cannot address unmeasured confounding. In our setting, plausible unmeasured (or coarsely measured) factors include posterior tibial slope, granularity of rotatory laxity severity within high-grade categories, precise tunnel position and graft fixation characteristics, extent of anterolateral/meniscocapsular pathology, rehabilitation adherence and training load, and surgeon-specific nuances not fully captured by the centre variable. If these were imbalanced, they could either attenuate or inflate the observed association (e.g., higher tibial slope or greater pre-injury activity intensity concentrated in the isolated ACLR group would bias towards benefit with LET). The absence of a significant LET×graft interaction and the consistency of direction across subgroups partly mitigate, without eliminating, this concern. Prospective randomised trials with detailed biomechanical and imaging covariates remain necessary.

4.5. Future Research

Ongoing STABILITY II and STABILITY Kids trials will refine risk-stratification algorithms. Randomised comparisons between LET and ALL reconstruction, and cost-effectiveness analyses, are warranted. Investigation into sex-specific responses and interaction with graft choice (quadriceps tendon) should also be prioritised.

5. Conclusions

In competitive athletes with high-grade rotatory laxity, adding a lateral extra-articular tenodesis to anatomic ACL reconstruction is associated with improved graft survival, lower residual pivot-shift and was linked to a faster criteria-based return to sport without increasing complications. With an NNT of eight to prevent one failure and negligible added morbidity, the modified Lemaire LET represents a valuable adjunct for high-risk knees undergoing modern accelerated rehabilitation. Wider adoption should be accompanied by surgeon training and objective RTS frameworks to optimise long-term joint health.